We propose to develop new therapeutic modalities for human tumors using as a model human malignancies heterotransplanted in nude mice. In particular, we propose: 1) to develop fast and reliable assays for testing the effectiveness of drugs and other therapeutic modalities (hormones, ionizing radiations and hyperthermia) against human tumors. In addition to improving the subcutaneous and intrarenal systems, we are investigating the intracerebral system which presents unique possibilities for studies of the blood-brain barrier permeability to certain drugs and the anterior chamber of the eye which offers the possibility to perform rapidly microtests of drug sensitivity on minimal amounts of tumors; 2) to study the predictivity of human tumors heterotransplants for clinical therapy by a) testing known antitumor agents or combinations of them against individual human tumors and comparing the effect of the same agent(s) against the same tumor in the nude and in the patient of origin; b) testing the statistical predictivity by means of tumor panels. We will develop panels of 14 tumors each for lung carcinoma, carcinoma of the uterus and carcinoma of the ovary (as we already did for colon and breast carcinomas and melanomas) and test nine known antitumor drugs against all 42 tumors; 3) to investigate the effect on human tumors of drugs which have a short biological half life and have to be continuously administered. We have developed a highly efficient system for the continuous infusion of drugs subcutaneously and intraperitoneally in nude mice. Using such a system, we are in the process of studying the effect of Thymidine and have started the study of hydroxyurea and of Ara C.